Validity and reliability of the 2-min step test in individuals with stroke and lower-limb musculoskeletal disorders

Abstract

IntroductionWe investigated the reliability and validity of the 2-min step test (2MST) for assessing the exercise endurance of individuals with stroke and lower-limb musculoskeletal disorders.Participants and methodsThe participants were 39 individuals with stroke and 42 with lower-limb musculoskeletal disorders (mainly hip fractures) from the convalescent rehabilitation wards of four hospitals. The concurrent validity and congruence between the 2MST and the 6-min walk test (6MWT) and construct validity by hypotheses testing, including mobility and lower limb muscle strength, were also confirmed. A subset of participants (stroke-group, n = 15; musculoskeletal-group, n = 19) underwent a retest 2MST for our evaluation of relative and absolute reliability using the intraclass correlation coefficient (ICC1,1) and Bland–Altman plot.ResultsBoth groups showed a moderate correlation between the 2MST and 6MWT (ρ = 0.55–0.60), but the congruence was not sufficient. The 6MWT was correlated with mobility in both groups and with muscle strength in the stroke group, whereas the 2MST did not show a significant correlation with mobility. The relative reliability was excellent in both groups (ICC1,1 > 0.9). In terms of absolute reliability, the width of the limit of agreement was 18.8% for the stroke group and 15.4% for the musculoskeletal group, relative to their respective sample means of 2MST. A fixed bias was identified in the stroke group, in which step counts increased by 6.5 steps upon retesting.DiscussionOur analyses revealed that the 2MST is a valid and reliable tool for assessing the exercise endurance of individuals with stroke or lower-limb musculoskeletal disorders. However, it is necessary to validate the absolute reliability observed herein by using a larger sample size. In addition, when assessing the exercise endurance of individuals with stroke, it may be necessary to consider the potential bias of an increased step count during retesting.